Project Summary/Abstract We propose to construct multi organ microphysiological systems (?Body-on-a-Chip? or BoaCs) from human and rat cells to use as a basis to understand species differences in response to exposure to drugs or chemicals in this new platform. The results will then be compared to clinical data, where available, and to archived in vivo animal data. This work will directly test whether such in vitro models can accurately reproduce species differences in response to known drugs. A preclinical model based on human cells that can accurately predict human response should lead to better decisions on whether exposure to a chemical or chemical mixture will be harmful to humans. An advantage of this in vitro approach, compared to standard in vitro systems (e.g. such as multiwell plates), is that the tissues can exchange metabolites and the dose dynamics in the body of both parental compounds and metabolites are better represented than when a single cell type is exposed to a bolus dose. Also, by comparing acute to chronic effects it will enable prediction on clinical trial success as well for determining PK of the compounds. In addition, the comparison of animal cells derived from iPSCs will enable the assessment of whether they can be substituted for primary animal cells. If successful, this could lead to stable cell sources for the animal models and reduce the number of animals needed for these studies. For this proposal we will build upon a four-organ model we recently published in Nature Scientific Reports (Oleaga, et al. 2016) which included model tissues for the liver, cardiac, skeletal muscle, and neuronal compartments that correctly predicted clinical response to five compounds. To construct a well defined system we will use a common serum free medium which mimics key features of blood. Hickman has developed microelectrode arrays and cantilever systems that are integrated on chip that allow for noninvasive electronic and mechanical readouts for not only acute but also chronic tests as well. To improve operability and enable a low volume system for eventual metabolite evaluation, we will use a pumpless system (Sung, et al. 210) and self contained devices. We will also utilize microfluidic analytical components for rapid and sensitive biomarker assessment. However, the number of biomarkers to be monitored for cell health and function will be greatly reduced in our systems from use of the function readouts. The system will be modeled by simulation using CFD to establish acceptable ranges for consumption of nutrients and drug metabolism as well as shear stress and to predict drug concentration profiles in the system to also enable PK/PD prediction capabilities. We believe that this technique will lead to more accurate and cost-effective assessment of the efficacy and toxicological potential of drugs chemicals or chemical mixtures and this approach will have a major impact on improving human health.